This is gated behind a new option in `podman system migrate`,
`--migrate-db`, or by a system restart being performed.
BoltDB support was removed in Podman 6, so we are certain that,
when we start Podman, a SQLite state is in use. However, if we
also detect a valid BoltDB state, we will attempt a migration.
Migration is performed by retrieving all volumes, pods, and
containers (in that order, to ensure there are no dependency
conflicts) from the Bolt database, when adding them to the SQLite
database. If there is a conflict - IE, a container exists in both
SQLite and Bolt - we skip migration for that object. The old DB
is then renamed so we do not try to migrate it again.
Our ability to test complex migration scenarios is limited, but
this should handle simple migrations easily.
This is a heavily adapted version of #27660 rebuilt to work with
Podman 6.0. Substantial changes were required to throw errors
when a BoltDB database is detected and no migration is being
performed. Firstly, for automatic on-reboot migrations, we need
to have a deferred error returned by getDBState (very early in
runtime initialization) that is only acted on much later (once we
know for certain a state refresh is/is not being performed).
The `system migrate --migrate-db` command was much more
problematic. Conceptually, it's not terrible - add a flag to the
runtime to suppress errors, set that flag only when calling the
`system migrate` command with `--migrate-db` - but it unveiled a
serious problem with how we do runtime init (special flags to the
runtime were being ignored because the image runtime set the
Libpod runtime first and had none of the proper handling) which
took a genuinely annoying amount of time to identify and fix.
This cannot be tested automatically, as the ability to create Bolt
databases has been entirely removed with Podman 6.
This also includes
|
||
|---|---|---|
| .. | ||
| artifact | ||
| common | ||
| completion | ||
| containers | ||
| diff | ||
| farm | ||
| generate | ||
| healthcheck | ||
| images | ||
| inspect | ||
| kube | ||
| machine | ||
| manifest | ||
| networks | ||
| parse | ||
| pods | ||
| quadlet | ||
| registry | ||
| secrets | ||
| system | ||
| utils | ||
| validate | ||
| volumes | ||
| auto-update.go | ||
| client.go | ||
| client_supported.go | ||
| client_unsupported.go | ||
| compose.go | ||
| compose_machine.go | ||
| compose_machine_unix.go | ||
| compose_machine_unsupported.go | ||
| compose_machine_windows.go | ||
| diff.go | ||
| early_init_darwin.go | ||
| early_init_linux.go | ||
| early_init_unsupported.go | ||
| inspect.go | ||
| login.go | ||
| logout.go | ||
| main.go | ||
| README.md | ||
| root.go | ||
| root_cgroups_linux.go | ||
| root_cgroups_unsupported.go | ||
| root_test.go | ||
| shell_completion_test.go | ||
| syslog_common.go | ||
| syslog_unsupported.go | ||
Podman CLI
The following is an example of how to add a new primary command (manifest) and a sub-command (inspect) to the Podman CLI.
This is example code, the production code has additional error checking and the business logic provided.
See items below for details on building, installing, contributing to Podman:
Adding a new command podman manifest
$ mkdir -p $GOPATH/src/github.com/containers/podman/cmd/podman/manifest
Create the file $GOPATH/src/github.com/containers/podman/cmd/podman/manifest/manifest.go
package manifest
import (
"go.podman.io/podman/v6/cmd/podman/registry"
"go.podman.io/podman/v6/cmd/podman/validate"
"go.podman.io/podman/v6/pkg/domain/entities"
"github.com/spf13/cobra"
)
var (
// podman _manifests_
manifestCmd = &cobra.Command{
Use: "manifest",
Short: "Manage manifests",
Args: cobra.ExactArgs(1),
Long: "Manage manifests",
Example: "podman manifest IMAGE",
TraverseChildren: true,
RunE: validate.SubCommandExists, // Report error if there is no sub command given
}
)
func init() {
// Subscribe command to podman
registry.Commands = append(registry.Commands, registry.CliCommand{
Command: manifestCmd,
})
}
To "wire" in the manifest command, edit the file $GOPATH/src/github.com/containers/podman/cmd/podman/main.go to add:
package main
import _ "go.podman.io/podman/v6/cmd/podman/manifest"
Adding a new sub command podman manifest list
Create the file $GOPATH/src/github.com/containers/podman/cmd/podman/manifest/inspect.go
package manifest
import (
"go.podman.io/podman/v6/cmd/podman/registry"
"go.podman.io/podman/v6/pkg/domain/entities"
"github.com/spf13/cobra"
)
var (
// podman manifests _inspect_
inspectCmd = &cobra.Command{
Use: "inspect IMAGE",
Short: "Display manifest from image",
Long: "Displays the low-level information on a manifest identified by image name or ID",
RunE: inspect,
Annotations: map[string]string{
// Add this annotation if this command cannot be run rootless
// registry.ParentNSRequired: "",
},
Example: "podman manifest inspect DEADBEEF",
}
)
func init() {
// Subscribe inspect sub command to manifest command
registry.Commands = append(registry.Commands, registry.CliCommand{
Command: inspectCmd,
// The parent command to proceed this command on the CLI
Parent: manifestCmd,
})
// This is where you would configure the cobra flags using inspectCmd.Flags()
}
// Business logic: cmd is inspectCmd, args is the positional arguments from os.Args
func inspect(cmd *cobra.Command, args []string) error {
// Business logic using registry.ImageEngine()
// Do not pull from libpod directly use the domain objects and types
return nil
}
Helper functions
The complete set can be found in the validate package, here are some examples:
cobra.Command{ Args: validate.NoArgs }used when the command does not accept errorscobra.Command{ Args: validate.IdOrLatestArgs }used to ensure either a list of ids given or the --latest flagcobra.Command{ RunE: validate.SubCommandExists }used to validate a subcommand given to a commandvalidate.ChoiceValueused to create apflag.Valuethat validate user input against a provided slice of values. For example:flags := cobraCommand.Flags() created := validate.ChoiceValue(&opts.Sort, "command", "created", "id", "image", "names", "runningfor", "size", "status") flags.Var(created, "sort", "Sort output by: "+created.Choices())
Adding CLI flags
When adding a new cli option that accepts a string array, there are two options to choose from: StringSlice() and StringArray().
They differ slightly in their behavior: StringSlice() allows the values to be comma separated so --opt v1,v2 --opt v3 results in
[]string{"v1", "v2", "v3"}, while StringArray() would result in []string{"v1,v2", "v3"}. Thus it is impossible to use values with comma in StringSlice(), which makes it unsuitable for flags that accept arbitrary values such as file paths as example. Also, because StringSlice() uses the csv lib to parse the values, it has special escaping rules for things like quotes, see https://github.com/containers/podman/issues/20064 for an example of how complicated things can get because of this.
Thus use StringSlice() only when the option accepts predefined values that do not contain special characters, for example --cap-add and --cap-drop are a good example for this. Using --cap-add NET_ADMIN,NET_RAW is equal to --cap-add NET_ADMIN --cap-add NET_RAW so it is better suited to save some typing for users.
When in doubt always choose StringArray() over StringSlice().